Open and Welcome Thread – July 2021 2021-07-03T19:53:07.048Z
Open and Welcome Thread – June 2021 2021-06-06T02:20:22.421Z
Attributions, Karma and better discoverability for wiki/tag features 2021-06-02T23:47:03.604Z
Open and Welcome Thread - May 2021 2021-05-03T07:58:03.130Z
2019 Review: Voting Results! 2021-02-01T03:10:19.284Z
Last day of voting for the 2019 review! 2021-01-26T00:46:35.426Z
The Great Karma Reckoning 2021-01-15T05:19:32.447Z
COVID-19: home stretch and fourth wave Q&A 2021-01-06T22:44:29.382Z
The medical test paradox: Can redesigning Bayes rule help? 2020-12-24T22:08:30.602Z
The LessWrong 2019 Review 2020-12-02T11:21:11.533Z
Sunday November 29th, 12:00PM (PT) — Andreas Stuhlmüller on Combining GPT-3 and Forecasting 2020-11-26T22:14:35.540Z
[Event] Ajeya's Timeline Report Reading Group #1 (Nov. 17, 6:30PM - 8:00PM PT) 2020-11-14T19:14:43.550Z
Sunday November 15th, 12:00PM (PT) — talks by Abram Demski, Daniel Kokotajlo and (maybe) more! 2020-11-13T00:53:17.126Z
Open & Welcome Thread – November 2020 2020-11-03T20:46:12.745Z
Sunday October 11th, 12:00PM (PT) — talks by Alex Zhu, Martin Sustrik and Steve Byrnes 2020-10-07T23:22:14.050Z
The new Editor 2020-09-23T02:25:53.914Z
AI Advantages [Gems from the Wiki] 2020-09-22T22:44:36.671Z
Sunday September 27, 12:00PM (PT) — talks by Alex Flint, Alex Zhu and more 2020-09-22T21:59:56.546Z
Gems from the Wiki: Do The Math, Then Burn The Math and Go With Your Gut 2020-09-17T22:41:24.097Z
Sunday September 20, 12:00PM (PT) — talks by Eric Rogstad, Daniel Kokotajlo and more 2020-09-17T00:27:47.735Z
Gems from the Wiki: Paranoid Debating 2020-09-15T03:51:10.453Z
Gems from the Wiki: Acausal Trade 2020-09-13T00:23:32.421Z
Notes on good judgement and how to develop it (80,000 Hours) 2020-09-12T17:51:27.174Z
How Much Computational Power Does It Take to Match the Human Brain? 2020-09-12T06:38:29.693Z
What's Wrong with Social Science and How to Fix It: Reflections After Reading 2578 Papers 2020-09-12T01:46:07.349Z
‘Ugh fields’, or why you can’t even bear to think about that task (Rob Wiblin) 2020-09-11T20:31:00.990Z
Sunday September 13, 12:00PM (PT) — talks by John Wentworth, Liron and more 2020-09-10T19:49:06.325Z
How To Fermi Model 2020-09-09T05:13:19.243Z
Conflict, the Rules of Engagement, and Professionalism 2020-09-05T05:04:16.081Z
Open & Welcome Thread - September 2020 2020-09-04T18:14:17.056Z
Sunday September 6, 12pm (PT) — Casual hanging out with the LessWrong community 2020-09-03T02:08:25.687Z
Open & Welcome Thread - August 2020 2020-08-06T06:16:50.337Z
Use resilience, instead of imprecision, to communicate uncertainty 2020-07-20T05:08:52.759Z
The New Frontpage Design & Opening Tag Creation! 2020-07-09T04:37:01.137Z
AI Research Considerations for Human Existential Safety (ARCHES) 2020-07-09T02:49:27.267Z
Open & Welcome Thread - July 2020 2020-07-02T22:41:35.440Z
Open & Welcome Thread - June 2020 2020-06-02T18:19:36.166Z
[U.S. Specific] Free money (~$5k-$30k) for Independent Contractors and grant recipients from U.S. government 2020-04-10T05:00:35.435Z
[Announcement] LessWrong will be down for ~1 hour on the evening of April 10th around 10PM PDT (5:00AM GMT) 2020-04-09T05:09:24.241Z
April Fools: Announcing LessWrong 3.0 – Now in VR! 2020-04-01T08:00:15.199Z
Rob Bensinger's COVID-19 overview 2020-03-28T21:47:31.684Z
Coronavirus Research Ideas for EAs 2020-03-27T22:10:35.767Z
March 25: Daily Coronavirus Updates 2020-03-27T04:32:18.530Z
March 24th: Daily Coronavirus Link Updates 2020-03-26T02:22:35.214Z
March 22nd & 23rd: Coronavirus Link Updates 2020-03-25T01:08:14.499Z
March 21st: Daily Coronavirus Links 2020-03-23T00:43:29.913Z
March 20th: Daily Coronavirus Links 2020-03-21T19:17:33.320Z
March 19th: Daily Coronavirus Links 2020-03-21T00:00:54.173Z
Sarah Constantin: Oxygen Supplementation 101 2020-03-20T01:00:16.453Z
March 18th: Daily Coronavirus Links 2020-03-19T22:20:27.217Z


Comment by habryka (habryka4) on Delta Strain: Fact Dump and Some Policy Takeaways · 2021-08-02T01:30:37.864Z · LW · GW

According to this specific calculation. He does say his all things considered view is like 5x the cost.

Comment by habryka (habryka4) on Delta Strain: Fact Dump and Some Policy Takeaways · 2021-08-01T19:46:15.490Z · LW · GW

The calculation would be pretty straightforward: 

3.5 days mean, let's assume another 2 days or so in terms of sickness, so 5.5 days lost in total for 1M microcovids. 

5.5 days = 130 hours

Which implies that ~7600 (1MM / 130) microcovids is ~1 hour of life lost.

Comment by habryka (habryka4) on Incorrect hypotheses point to correct observations · 2021-07-31T05:55:09.717Z · LW · GW

Shouldn't be too hard, but seemed worth testing whether it was working in a low-tech solution before we build a thing. My guess is we will add this more properly relatively soon, if we keep doing it.

Comment by habryka (habryka4) on Incorrect hypotheses point to correct observations · 2021-07-30T23:33:01.270Z · LW · GW

One of the reasons for this is that curated posts get emailed out to a few thousand people and that email doesn't include any comments, and frequently others have asked us for context on why we curated something when they get the email. So this way the curation notice gets to be included in the email.

Comment by habryka (habryka4) on DeepMind: Generally capable agents emerge from open-ended play · 2021-07-30T03:52:55.451Z · LW · GW

Sometimes Eliezer's posts are amazing, sometimes they are only OK. Saturating the karma system would get rid of the distinction between those, and also break any sorting we might want to do on the frontpage based on karma.

Comment by habryka (habryka4) on DeepMind: Generally capable agents emerge from open-ended play · 2021-07-29T20:19:36.978Z · LW · GW

Yep, the author's vote-strength is used for the initial vote. So comments by users with more karma can start out with 2 points, as opposed to the default 1, and posts by users with more karma can start with up to 10 points.

Comment by habryka (habryka4) on Working hurts less than procrastinating, we fear the twinge of starting · 2021-07-29T17:19:19.907Z · LW · GW

Mod note: This... is not an appropriate level of response for the forum. The above comment does not indicate that it's author "believes every article written about everything". Take this as a mod warning, more comments like this and we might ban you.

Comment by habryka (habryka4) on DeepMind: Generally capable agents emerge from open-ended play · 2021-07-29T04:48:44.582Z · LW · GW

I think voting should happen primarily based on quality of contribution, not norm violation. For norm violations you can report it to the mods. I think it's pretty reasonable for some comments that don't violate any norms but just seem generically low quality to get downvoted (and indeed, most of the internet is full of low-quality contributions that make discussions frustrating and hard to interact with, without violating any clear norms, and our downvote system holds them at least somewhat at bay). 

Comment by habryka (habryka4) on DeepMind: Generally capable agents emerge from open-ended play · 2021-07-29T04:47:26.236Z · LW · GW

There is no automatic voting routine. But all comments start with a single small-upvote by their author, and posts start with a single strong-upvote by their author.

Comment by habryka (habryka4) on Covid 7/22: Error Correction · 2021-07-23T20:07:28.529Z · LW · GW


Comment by habryka (habryka4) on ($1000 bounty) How effective are marginal vaccine doses against the covid delta variant? · 2021-07-22T23:24:21.043Z · LW · GW

Huh, that's really interesting. Do you have any sources around this? That sure could introduce some big shifts in the EV.

Comment by habryka (habryka4) on Covid 7/22: Error Correction · 2021-07-22T20:36:38.648Z · LW · GW

See also this question, which asks the same question and comes with a $1000 bounty for the best answers: 

Comment by habryka (habryka4) on [AN #131]: Formalizing the argument of ignored attributes in a utility function · 2021-07-17T19:18:20.662Z · LW · GW

Oops, indeed. Fixed.

Comment by habryka (habryka4) on How to Sleep Better · 2021-07-17T03:47:40.306Z · LW · GW

Huh, interesting, I haven't heard of this. Do you have a source handy?

Comment by habryka (habryka4) on Matthew Barnett's Shortform · 2021-07-16T23:26:47.218Z · LW · GW

I usually associate things like "being evil" more with something like "part of my payoff matrix has a negative coefficient on your payoff matrix". I.e. actively wanting to hurt people and taking inherent interest in making them worse off. Selfishness feels pretty different from being evil emotionally, at least to me.

Comment by habryka (habryka4) on AlphaFold 2 paper released: "Highly accurate protein structure prediction with AlphaFold", Jumper et al 2021 · 2021-07-15T22:08:52.642Z · LW · GW

Parallel HN thread: 

Comment by habryka (habryka4) on What would it look like if it looked like AGI was very near? · 2021-07-14T06:54:56.577Z · LW · GW

Oh, well, I definitely didn't get that reference. But that definitely makes me more sympathetic to the title.

Comment by habryka (habryka4) on What would it look like if it looked like AGI was very near? · 2021-07-12T23:24:39.722Z · LW · GW

Would the title of this just be better as "What would it look like if AGI was near?". I feel a bit confused what the additional "if it looked like" clause is doing.

Comment by habryka (habryka4) on Intermittent Distillations #4: Semiconductors, Economics, Intelligence, and Technological Progress. · 2021-07-09T20:31:08.902Z · LW · GW

I also found these very valuable! I wonder whether a better title might help more people see how great these are, but not sure.

Comment by habryka (habryka4) on Improving capital gains taxes · 2021-07-09T20:12:03.665Z · LW · GW

Mod note: For critiques like this, please give more context or background. Disagreement and critiques are really useful, but they are very hard to engage with when they give this little context. In my experience, comments like this can cause quite a bit of stress for both readers and authors, and I would really like to avoid that on LessWrong, so try to outline at least a basic argument for why you believe something when you disagree, or alternatively add a meta-comment of the form "I don't have time to write up my reasons for believing this, but I am overall pretty confident that this would reduce tax revenue. Sorry that I don't have time to write up the arguments for this, but I think that's a major obstacle to this being a good idea, and it seemed good to voice my take even if I can't give more context", or something like that.

Comment by habryka (habryka4) on Announcing My Free Online Course "Original Seeing With a Focus On Life" · 2021-07-09T04:29:24.160Z · LW · GW

Oh, yeah. I tried to scroll, but on a desktop clicking and dragging wasn't something I considered.

Comment by habryka (habryka4) on Announcing My Free Online Course "Original Seeing With a Focus On Life" · 2021-07-08T05:40:05.154Z · LW · GW

I was unable to read all the quotes. At least one of them was cut off on the right side: 

I also couldn't find any way to scroll to make the quote visible.

Comment by habryka (habryka4) on Measuring hardware overhang · 2021-07-08T05:30:08.741Z · LW · GW

Replaced the image in the post with this image.

Comment by habryka (habryka4) on Habryka's Shortform Feed · 2021-07-06T20:54:58.280Z · LW · GW

This FB post by Matt Bell on the Delta Variant helped me orient a good amount:

As has been the case for almost the entire pandemic, we can predict the future by looking at the present. Let’s tackle the question of “Should I worry about the Delta variant?” There’s now enough data out of Israel and the UK to get a good picture of this, as nearly all cases in Israel and the UK for the last few weeks have been the Delta variant. [1] Israel was until recently the most-vaccinated major country in the world, and is a good analog to the US because they’ve almost entirely used mRNA vaccines.

- If you’re fully vaccinated and aren’t in a high risk group, the Delta variant looks like it might be “just the flu”. There are some scary headlines going around, like “Half of new cases in Israel are among vaccinated people”, but they’re misleading for a couple of reasons. First, since Israel has vaccinated over 80% of the eligible population, the mRNA vaccine still is 1-((0.5/0.8)/(0.5/0.2)) = 75% effective against infection with the Delta variant. Furthermore, the efficacy of the mRNA vaccine is still very high ( > 90%) against hospitalization or death from the Delta variant. Thus, you might still catch Delta if you’re vaccinated, but it will be more like a regular flu if you do. J&J likely has a similar performance in terms of reduced hospitalizations and deaths, as the UK primarily vaccinated its citizens with the AstraZeneca vaccine, which is basically a crappier version of J&J, and is still seeing a 90+% reduction in hospitalizations and deaths among the vaccinated population. [2]


Comment by habryka (habryka4) on How credible is the theory that COVID19 escaped from a Wuhan Lab? · 2021-07-06T18:31:56.266Z · LW · GW

Do you have a new probability?

Comment by habryka (habryka4) on Reframing the evolutionary benefit of sex · 2021-07-06T18:30:20.875Z · LW · GW

Ah, that section is a random sample of posts that you haven't read, weighted slightly exponentially by their karma. So any frontpage post can show up in that section.

Comment by habryka (habryka4) on Reframing the evolutionary benefit of sex · 2021-07-05T22:22:04.713Z · LW · GW

I'm not sure why this is pinned.

Not sure what you mean by this?

Comment by habryka4 on [deleted post] 2021-07-04T05:48:18.945Z

Seems great to me. I will revisit it in a month or so and see whether it has reached critical mass.

Comment by habryka4 on [deleted post] 2021-06-28T20:32:05.193Z

I think we should just delete this tag for now. There are three posts, and two aren't even a good fit.

Comment by habryka (habryka4) on johnswentworth's Shortform · 2021-06-25T20:00:30.969Z · LW · GW

Huh, amusing. We do ship a font that has nothing but the greek letter set in it, because people use greek unicode symbols all the time and our primary font doesn't support that character set. So my guess is that's where Google gets confused.

Comment by habryka (habryka4) on On the limits of idealized values · 2021-06-24T20:22:10.214Z · LW · GW

Promoted to curated: I really liked this post. I've had some thoughts along similar lines for a while, and this post clarified a bunch of them in much better ways than I have succeeded at so far. It also seems like a pretty important topic. Thank you for writing this!

Comment by habryka (habryka4) on Alex Turner's Research, Comprehensive Information Gathering · 2021-06-23T17:50:49.823Z · LW · GW

Minor meta feedback: I think it's better to put the "Comprehensive Information Gathering" part of the title at the end, if you want to have many of these. That makes it much easier to see differences in the title and skim a list of them.

Comment by habryka (habryka4) on Less Wrong needs footnotes · 2021-06-22T16:22:00.738Z · LW · GW

Yep, markdown has footnotes. It's been on my to-do list for a while to add footnotes to the WYSIWYG editor, it was just a bit of an annoying UI challenge.

Comment by habryka (habryka4) on Why did no LessWrong discourse on gain of function research develop in 2013/2014? · 2021-06-20T03:29:27.196Z · LW · GW

Some relevant links: 

Don't think they prove anything, but seem useful references.

Comment by habryka (habryka4) on Why did no LessWrong discourse on gain of function research develop in 2013/2014? · 2021-06-19T00:17:35.771Z · LW · GW

I feel like this just happened? There was a good amount of articles written about this, see for example this article by the Global Priorities Project on GoF research: 

I also remember a number of other articles by people working on biorisk, but would have to dig them up. But overall I had a sense there was both a bunch of funding and a bunch of advocacy and a bunch of research on this topic.

Comment by habryka (habryka4) on What are examples of the opposite of perverse incentives? · 2021-06-18T19:41:21.602Z · LW · GW

I edited it

Comment by habryka (habryka4) on Open and Welcome Thread – June 2021 · 2021-06-18T19:40:00.066Z · LW · GW

Having someone else proofread it, or just generally give style feedback, is often quite valuable. You can post in the Open Thread with a link to a Google Doc, or you can ping me on Intercom, and I can take a look and suggest any obvious improvements that stand out.

Also, there is value to independently posting a link to your essay via other non-LW channels. E.g. sharing it on Facebook or Discord or wherever you might find others interested in it.

Comment by habryka (habryka4) on Are bread crusts healthier? · 2021-06-18T18:56:38.190Z · LW · GW

I am from germany, and I don't really agree with this assessment of the crust being the tastiest part. People definitely told me that many times, but I think it does just seem straightforwardly wrong by my own taste. I do still usually eat the crust though.

Comment by habryka (habryka4) on [AN #152]: How we’ve overestimated few-shot learning capabilities · 2021-06-16T22:45:50.244Z · LW · GW

The newsletter is back! I missed these! Glad to have these back.

Comment by habryka (habryka4) on Open and Welcome Thread – June 2021 · 2021-06-14T05:27:23.310Z · LW · GW


Comment by habryka (habryka4) on Open and Welcome Thread – June 2021 · 2021-06-13T04:59:19.998Z · LW · GW

There is a private FB event I am invited to and I can invite you to. I don't think we're friends on FB, so if you have FB, you can add me, and then I can invite you: 

Comment by habryka (habryka4) on Rogue AGI Embodies Valuable Intellectual Property · 2021-06-11T05:30:09.446Z · LW · GW

Promoted to curated: I've had a number of disagreements with a perspective on AI that generates arguments like the above, which takes something like "ownership of material resources" as a really fundamental unit of analysis, and I feel like this post has both helped me get a better grasp on that paradigm of thinking, and also helped me get a bit of a better sense of what feels off to me, and I have a feeling this post will be useful in bridging that gap eventually. 

Comment by habryka (habryka4) on The dumbest kid in the world (joke) · 2021-06-11T05:06:36.673Z · LW · GW

It's a bit janky, but I've gotten it to work by selecting the text between the two vote symbols. Definitely not as accessible as I would like though.

Comment by habryka (habryka4) on ML is now automating parts of chip R&D. How big a deal is this? · 2021-06-11T05:03:52.900Z · LW · GW

This is a great comment! Thank you for writing it! 

Comment by habryka (habryka4) on Habryka's Shortform Feed · 2021-06-08T04:16:33.927Z · LW · GW

This seems like potentially a big deal:

> Troubling—the worst variant to date, the #DeltaVariant is now the new fastest growing variant in US. This is the so-called “Indian” variant #B16172 that is ravaging the UK despite high vaccinations because it has immune evasion properties. Here is why it’s trouble—Thread. #COVID19

Comment by habryka (habryka4) on Power dynamics as a blind spot or blurry spot in our collective world-modeling, especially around AI · 2021-06-08T03:42:16.397Z · LW · GW

This also summarizes a lot of my take on this position. Thank you!

Comment by habryka (habryka4) on The Alignment Forum should have more transparent membership standards · 2021-06-07T02:11:57.569Z · LW · GW

I do think "non-member AIAF" account is a useful abstraction, just one we haven't leveraged very much. But, for example, it does allow you to respond to comments on your post on the AIAF, even if you are not a member. So the signup does make sense.

Comment by habryka (habryka4) on The Alignment Forum should have more transparent membership standards · 2021-06-06T19:45:19.010Z · LW · GW

My general philosophy for things like this is "do it in-house for a while so you understand what kinds of problems come up and make sure you have a good experience. After you really have it down maybe consider outsourcing it, or requesting volunteer labor for it." 

So I think eventually asking for a more crowdsourced solution seems reasonable, though I think that would come a few months after. 

Comment by habryka (habryka4) on Open and Welcome Thread - May 2021 · 2021-06-06T02:21:35.530Z · LW · GW

I literally 2 minutes ago created the June Open thread for this year and pinned that one. So if I were you I would probably repost this there instead of here: 

Comment by habryka (habryka4) on The Alignment Forum should have more transparent membership standards · 2021-06-06T01:48:55.230Z · LW · GW

Yeah, I was proposing something like this in this comment response to Peter.